lib/lyx2lyx/lyxconvert_218.py

   1 # This file is part of lyx2lyx
   2 # Copyright (C) 2002 Dekel Tsur <dekel@lyx.org>
   3 #
   4 # This program is free software; you can redistribute it and/or
   5 # modify it under the terms of the GNU General Public License
   6 # as published by the Free Software Foundation; either version 2
   7 # of the License, or (at your option) any later version.
   8 #
   9 # This program is distributed in the hope that it will be useful,
  10 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  11 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12 # GNU General Public License for more details.
  13 #
  14 # You should have received a copy of the GNU General Public License
  15 # along with this program; if not, write to the Free Software
  16 # Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
  17
  18
  19 import sys,string,re
  20 from parser_tools import *
  21
  22 floats = {
  23     "footnote": ["\\begin_inset Foot",
  24                  "collapsed true"],
  25     "margin":   ["\\begin_inset Marginal",
  26                  "collapsed true"],
  27     "fig":      ["\\begin_inset Float figure",
  28                  "wide false",
  29                  "collapsed false"],
  30     "tab":      ["\\begin_inset Float table",
  31                  "wide false",
  32                  "collapsed false"],
  33     "alg":      ["\\begin_inset Float algorithm",
  34                  "wide false",
  35                  "collapsed false"],
  36     "wide-fig": ["\\begin_inset Float figure",
  37                  "wide true",
  38                  "collapsed false"],
  39     "wide-tab": ["\\begin_inset Float table",
  40                  "wide true",
  41                  "collapsed false"]
  42 }
  43
  44 font_tokens = ["\\family", "\\series", "\\shape", "\\size", "\\emph",
  45                "\\bar", "\\noun", "\\color", "\\lang", "\\latex"]
  46
  47 #
  48 # Change \begin_float .. \end_float into \begin_inset Float .. \end_inset
  49 #
  50
  51 pextra_type3_rexp = re.compile(r".*\\pextra_type\s+3")
  52 pextra_rexp = re.compile(r"\\pextra_type\s+(\S+)"+\
  53                          r"(\s+\\pextra_alignment\s+(\S+))?"+\
  54                          r"(\s+\\pextra_hfill\s+(\S+))?"+\
  55                          r"(\s+\\pextra_start_minipage\s+(\S+))?"+\
  56                          r"(\s+(\\pextra_widthp?)\s+(\S*))?")
  57
  58 def get_width(mo):
  59     if mo.group(10):
  60         if mo.group(9) == "\\pextra_widthp":
  61             return mo.group(10)+"col%"
  62         else:
  63             return mo.group(10)
  64     else:
  65         return "100col%"
  66
  67 def remove_oldfloat(lines, language):
  68     i = 0
  69     while 1:
  70         i = find_token(lines, "\\begin_float", i)
  71         if i == -1:
  72             break
  73         # There are no nested floats, so finding the end of the float is simple
  74         j = find_token(lines, "\\end_float", i+1)
  75
  76         floattype = string.split(lines[i])[1]
  77         if not floats.has_key(floattype):
  78             sys.stderr.write("Error! Unknown float type "+floattype+"\n")
  79             floattype = "fig"
  80
  81         # skip \end_deeper tokens
  82         i2 = i+1
  83         while check_token(lines[i2], "\\end_deeper"):
  84             i2 = i2+1
  85         if i2 > i+1:
  86             j2 = get_next_paragraph(lines, j+1)
  87             lines[j2:j2] = ["\\end_deeper "]*(i2-(i+1))
  88
  89         new = floats[floattype]+[""]
  90
  91         # Check if the float is floatingfigure
  92         k = find_re(lines, pextra_type3_rexp, i, j)
  93         if k != -1:
  94             mo = pextra_rexp.search(lines[k])
  95             width = get_width(mo)
  96             lines[k] = re.sub(pextra_rexp, "", lines[k])
  97             new = ["\\begin_inset Wrap figure",
  98                    'width "%s"' % width,
  99                    "collapsed false",
 100                    ""]
 101
 102         new = new+lines[i2:j]+["\\end_inset ", ""]
 103
 104         # After a float, all font attributes are reseted.
 105         # We need to output '\foo default' for every attribute foo
 106         # whose value is not default before the float.
 107         # The check here is not accurate, but it doesn't matter
 108         # as extra '\foo default' commands are ignored.
 109         # In fact, it might be safer to output '\foo default' for all
 110         # font attributes.
 111         k = get_paragraph(lines, i)
 112         flag = 0
 113         for token in font_tokens:
 114             if find_token(lines, token, k, i) != -1:
 115                 if not flag:
 116                     # This is not necessary, but we want the output to be
 117                     # as similar as posible to the lyx format
 118                     flag = 1
 119                     new.append("")
 120                 if token == "\\lang":
 121                     new.append(token+" "+language)
 122                 else:
 123                     new.append(token+" default ")
 124
 125         lines[i:j+1] = new
 126         i = i+1
 127
 128 pextra_type2_rexp = re.compile(r".*\\pextra_type\s+2")
 129 pextra_type2_rexp2 = re.compile(r".*(\\layout|\\pextra_type\s+2)")
 130
 131 def remove_oldminipage(lines):
 132     i = 0
 133     flag = 0
 134     while 1:
 135         i = find_re(lines, pextra_type2_rexp, i)
 136         if i == -1:
 137             break
 138
 139         mo = pextra_rexp.search(lines[i])
 140         position = mo.group(3)
 141         hfill = mo.group(5)
 142         width = get_width(mo)
 143         lines[i] = re.sub(pextra_rexp, "", lines[i])
 144
 145         start = ["\\begin_inset Minipage",
 146                  "position " + position,
 147                  "inner_position 0",
 148                  'height "0pt"',
 149                  'width "%s"' % width,
 150                  "collapsed false"
 151                  ]
 152         if flag:
 153             flag = 0
 154             if hfill:
 155                 start = ["","\hfill",""]+start
 156         else:
 157             start = ["\\layout Standard"] + start
 158
 159         j = find_token_backwards(lines,"\\layout", i-1)
 160         j0 = j
 161
 162         j = find_tokens(lines, ["\\layout", "\\end_float"], i+1)
 163         # j can be -1
 164
 165         count = 0
 166         while 1:
 167             # collect more paragraphs to the minipage
 168             count = count+1
 169             if j == -1 or not check_token(lines[j], "\\layout"):
 170                 break
 171             i = find_re(lines, pextra_type2_rexp2, j+1)
 172             if i == -1:
 173                 break
 174             mo = pextra_rexp.search(lines[i])
 175             if not mo:
 176                 break
 177             if mo.group(7) == "1":
 178                 flag = 1
 179                 break
 180             lines[i] = re.sub(pextra_rexp, "", lines[i])
 181             j = find_tokens(lines, ["\\layout", "\\end_float"], i+1)
 182
 183         mid = lines[j0:j]
 184         end = ["\\end_inset "]
 185
 186         lines[j0:j] = start+mid+end
 187         i = i+1
 188
 189 def is_empty(lines):
 190     return filter(is_nonempty_line, lines) == []
 191
 192 font_rexp =  re.compile(r"\\(family|series|shape|size|emph|numeric|bar|noun)")
 193 ert_rexp = re.compile(r"\\begin_inset|.*\\SpecialChar")
 194 spchar_rexp = re.compile(r"(.*)(\\SpecialChar.*)")
 195 ert_begin = ["\\begin_inset ERT",
 196              "status Collapsed",
 197              "",
 198              "\\layout Standard"]
 199
 200 def remove_oldert(lines):
 201     i = 0
 202     while 1:
 203         i = find_tokens(lines, ["\\latex latex", "\\layout LaTeX"], i)
 204         if i == -1:
 205             break
 206         j = i+1
 207         while 1:
 208             j = find_tokens(lines, ["\\latex default", "\\begin_inset", "\\layout", "\\end_float", "\\the_end"],
 209                             j)
 210             if check_token(lines[j], "\\begin_inset"):
 211                 j = find_end_of_inset(lines, j)
 212             else:
 213                 break
 214
 215         if check_token(lines[j], "\\layout"):
 216             while j-1 >= 0 and check_token(lines[j-1], "\\begin_deeper"):
 217                 j = j-1
 218
 219         # We need to remove insets, special chars & font commands from ERT text
 220         new = []
 221         new2 = []
 222         if check_token(lines[i], "\\layout LaTeX"):
 223             new = ["\layout Standard", "", ""]
 224             # We have a problem with classes in which Standard is not the default layout!
 225
 226         k = i+1
 227         while 1:
 228             k2 = find_re(lines, ert_rexp, k, j)
 229             inset = specialchar = 0
 230             if k2 == -1:
 231                 k2 = j
 232             elif check_token(lines[k2], "\\begin_inset"):
 233                 inset = 1
 234             else:
 235                 specialchar = 1
 236                 mo = spchar_rexp.match(lines[k2])
 237                 lines[k2] = mo.group(1)
 238                 specialchar_str = mo.group(2)
 239                 k2 = k2+1
 240
 241             tmp = []
 242             for line in lines[k:k2]:
 243                 if font_rexp.match(line):
 244                     if new2 == []:
 245                         # This is not necessary, but we want the output to be
 246                         # as similar as posible to the lyx format
 247                         new2 = [""]
 248                     new2.append(line)
 249                 elif not check_token(line, "\\latex"):
 250                     tmp.append(line)
 251
 252             if is_empty(tmp):
 253                 if filter(lambda x:x != "", tmp) != []:
 254                     if new == []:
 255                         # This is not necessary, but we want the output to be
 256                         # as similar as posible to the lyx format
 257                         lines[i-1] = lines[i-1]+" "
 258                     else:
 259                         new = new+[" "]
 260             else:
 261                 new = new+ert_begin+tmp+["\\end_inset ", ""]
 262
 263             if inset:
 264                 k3 = find_end_of_inset(lines, k2)
 265                 new = new+[""]+lines[k2:k3+1]+[""] # Put an empty line after \end_inset
 266                 k = k3+1
 267                 # Skip the empty line after \end_inset
 268                 if not is_nonempty_line(lines[k]):
 269                     k = k+1
 270                     new.append("")
 271             elif specialchar:
 272                 if new == []:
 273                     # This is not necessary, but we want the output to be
 274                     # as similar as posible to the lyx format
 275                     lines[i-1] = lines[i-1]+specialchar_str
 276                     new = [""]
 277                 else:
 278                     new = new+[specialchar_str, ""]
 279                 k = k2
 280             else:
 281                 break
 282
 283         new = new+new2
 284         if not check_token(lines[j], "\\latex "):
 285             new = new+[""]+[lines[j]]
 286         lines[i:j+1] = new
 287         i = i+1
 288
 289     i = 0
 290     while 1:
 291         i = find_token(lines, "\\latex ", i)
 292         if i == -1:
 293             break
 294         del lines[i]
 295
 296
 297 def remove_oldertinset(lines):
 298     i = 0
 299     while 1:
 300         i = find_token(lines, "\\begin_inset ERT", i)
 301         if i == -1:
 302             break
 303         j = find_end_of_inset(lines, i)
 304         k = find_token(lines, "\\layout", i+1)
 305         l = get_paragraph(lines, i)
 306         if lines[k] == lines[l]: # same layout
 307             k = k+1
 308         new = lines[k:j]
 309         lines[i:j+1] = new
 310         i = i+1
 311
 312 def is_ert_paragraph(lines, i):
 313     i = find_nonempty_line(lines, i+1)
 314     if not check_token(lines[i], "\\begin_inset ERT"):
 315         return 0
 316     j = find_end_of_inset(lines, i)
 317     k = find_nonempty_line(lines, j+1)
 318     return check_token(lines[k], "\\layout")
 319
 320 def combine_ert(lines):
 321     i = 0
 322     while 1:
 323         i = find_token(lines, "\\begin_inset ERT", i)
 324         if i == -1:
 325             break
 326         j = find_token_backwards(lines,"\\layout", i-1)
 327         count = 0
 328         text = []
 329         while is_ert_paragraph(lines, j):
 330
 331             count = count+1
 332             i2 = find_token(lines, "\\layout", j+1)
 333             k = find_token(lines, "\\end_inset", i2+1)
 334             text = text+lines[i2:k]
 335             j = find_token(lines, "\\layout", k+1)
 336             if j == -1:
 337                 break
 338
 339         if count >= 2:
 340             j = find_token(lines, "\\layout", i+1)
 341             lines[j:k] = text
 342
 343         i = i+1
 344
 345 oldunits = ["pt", "cm", "in", "text%", "col%"]
 346
 347 def get_length(lines, name, start, end):
 348     i = find_token(lines, name, start, end)
 349     if i == -1:
 350         return ""
 351     x = string.split(lines[i])
 352     return x[2]+oldunits[int(x[1])]
 353
 354 def write_attribute(x, token, value):
 355     if value != "":
 356         x.append("\t"+token+" "+value)
 357
 358 def remove_figinset(lines):
 359     i = 0
 360     while 1:
 361         i = find_token(lines, "\\begin_inset Figure", i)
 362         if i == -1:
 363             break
 364         j = find_end_of_inset(lines, i)
 365
 366         lyxwidth = string.split(lines[i])[3]+"pt"
 367         lyxheight = string.split(lines[i])[4]+"pt"
 368
 369         filename = get_value(lines, "file", i+1, j)
 370
 371         width = get_length(lines, "width", i+1, j)
 372         # what does width=5 mean ?
 373         height = get_length(lines, "height", i+1, j)
 374         rotateAngle = get_value(lines, "angle", i+1, j)
 375         if width == "" and height == "":
 376             size_type = "0"
 377         else:
 378             size_type = "1"
 379
 380         flags = get_value(lines, "flags", i+1, j)
 381         x = int(flags)%4
 382         if x == 1:
 383             display = "monochrome"
 384         elif x == 2:
 385             display = "gray"
 386         else:
 387             display = "color"
 388
 389         subcaptionText = get_value(lines, "subcaption", i+1, j)
 390         if subcaptionText != "":
 391             subcaptionText = '"'+subcaptionText+'"'
 392         k = find_token(lines, "subfigure", i+1,j)
 393         if k == -1:
 394             subcaption = 0
 395         else:
 396             subcaption = 1
 397
 398         new = ["\\begin_inset Graphics FormatVersion 1"]
 399         write_attribute(new, "filename", filename)
 400         write_attribute(new, "display", display)
 401         if subcaption:
 402             new.append("\tsubcaption")
 403         write_attribute(new, "subcaptionText", subcaptionText)
 404         write_attribute(new, "size_type", size_type)
 405         write_attribute(new, "width", width)
 406         write_attribute(new, "height", height)
 407         if rotateAngle != "":
 408             new.append("\trotate")
 409             write_attribute(new, "rotateAngle", rotateAngle)
 410         write_attribute(new, "rotateOrigin", "leftBaseline")
 411         write_attribute(new, "lyxsize_type", "1")
 412         write_attribute(new, "lyxwidth", lyxwidth)
 413         write_attribute(new, "lyxheight", lyxheight)
 414         new = new + ["\end_inset"]
 415         lines[i:j+1] = new
 416
 417 attr_re = re.compile(r' \w*="(false|0|)"')
 418 line_re = re.compile(r'<(features|column|row|cell)')
 419
 420 def update_tabular(lines):
 421     i = 0
 422     while 1:
 423         i = find_token(lines, '\\begin_inset  Tabular', i)
 424         if i == -1:
 425             break
 426
 427         for k in get_tabular_lines(lines, i):
 428             if check_token(lines[k], "<lyxtabular"):
 429                 lines[k] = string.replace(lines[k], 'version="2"', 'version="3"')
 430             elif check_token(lines[k], "<column"):
 431                 lines[k] = string.replace(lines[k], 'width=""', 'width="0pt"')
 432
 433             if line_re.match(lines[k]):
 434                 lines[k] = re.sub(attr_re, "", lines[k])
 435
 436         i = i+1
 437
 438 def change_preamble(lines):
 439     i = find_token(lines, "\\use_amsmath", 0)
 440     if i == -1:
 441         return
 442     lines[i+1:i+1] = ["\\use_natbib 0",
 443                       "\use_numerical_citations 0"]
 444
 445 def convert(header, body):
 446     language = get_value(header, "\\language", 0)
 447     if language == "":
 448         language = "english"
 449
 450     change_preamble(header)
 451     update_tabular(body)
 452     remove_oldminipage(body)
 453     remove_oldfloat(body, language)
 454     remove_figinset(body)
 455     remove_oldertinset(body)
 456     remove_oldert(body)
 457     combine_ert(body)
 458
 459 if __name__ == "__main__":
 460     pass